Reduced Surveillance Possible in Some Polyp Patients

August 1, 1997

NEW ORLEANS--Colorectal cancer surveillance intervals can be lengthened for some patients following initial polypectomy, two large studies suggest.

NEW ORLEANS--Colorectal cancer surveillance intervals can be lengthenedfor some patients following initial polypectomy, two large studies suggest.

Six-year follow-up data from the National Polyp Study showed that newlydiagnosed patients with three or more adenomas at initial colonoscopy wereat high risk for future polyps and should be re-examined at three years.But surveillance for patients with a single adenoma at baseline could beextended to at least six years, Ann G. Zauber, PhD, of Memorial Sloan-KetteringCancer Center, reported at the American Society of Preventive Oncologyannual meeting.

Current surveillance colonoscopy recommendations are not cost-effective,several speakers said at the meeting. It takes 320 colonoscopies to identifyone colo-rectal cancer. But 3% of adenomas found at 3-year follow-up colonoscopyare advanced and 0.6% are invasive cancer.

"This suggests that routine follow-up as currently recommendedoccurs too late for some patients but too early for most patients. Thesurveillance interval should be individualized, if possible," saidWei Zheng, MD, PhD, of the University of Minnesota, Minneapolis.

Dr. Zheng was the principal investigator for a large study of 1,490adenoma patients in Zheijiang, China, which found that risk of metachronousadenomas, especially advanced neoplasms, is closely related to the pathologiccharacteristics of the initial adenomas.

National Polyp Study Results

In the National Polyp Study mentioned above, 337 patients underwenttwo or three surveillance colonoscopies by the end of six years. Researcherswere looking for baseline characteristics that might predict risk for futuredisease.

At follow-up, 8% of patients had adenomas with advanced pathology (largerthan 1 cm, high-grade dysplasia, or infiltrating cancer); 42% had otheradenomas (1 cm or smaller, no high-grade dysplasia, no infiltration); and49% had no adenomas detected.

The greatest predictor of risk was number of adenomas at baseline: threeor more polyps carried an odds ratio of 15.7 for adenomas with advancedpathology. Fifteen (20%) of 74 patients with three or more adenomas atbaseline had advanced adenomas, compared with 7 (4%) of 187 patients witha single baseline adenoma.

Patients with a family history of colo-rectal cancer and who were age60 years or older at initial diagnosis were at an increased risk of 7.3,Dr. Zauber said.

Of the 28 patients with advanced adenomas at follow-up, 20 fell intothe high-risk category, either because of having multiple adenomas presentinitially or being age 60 or older with a family history. Five of theseadenomas were malignant, and the others were large adenomas not yet cancerous.

"This constitutes a relatively good baseline classification forpatients who need exams every three years," Dr. Zauber said. "Onthe other hand, patients with one or two adenomas at baseline, no parentalhistory of colorectal cancer, or initial diagnosis under age 60 can gosix or more years before surveillance."

Chinese Results

In the Chinese cohort, 280 of the 1,490 patients aged 30 and older developedadenomas in the distal colon and rectum in the 16-year follow-up period.Patients with large adenomas with severe dysplasia were found to have avery high risk (a 37-fold increase) of advanced recurrent adenomas in thefuture. These patients need close surveillance, Dr. Zheng said.

A 14-fold relative risk of advanced recurrent adenomas was found forpatients whose initial adenomas contained a high degree of dysplasia. Villous/tubulovillousadenomas (as opposed to tubular) carried an 8.0 relative risk, and patientswith baseline adenomas larger than 1.0 cm had a relative risk of 4.3